Pleural effusions in HIV-infected patients

INTRODUCTION — Pleural effusions are frequently encountered in patients with acquired immunodeficiency syndrome (AIDS). The prevalence of pleural effusion in hospitalized patients with AIDS varies from 7 to 27 percent [1-4]. The variations in prevalence may be explained by differences in the clinical stage of AIDS and the techniques used in detecting pleural effusion.

The majority of pleural effusions in patients with HIV infection are caused by infections; however, approximately a third are due to noninfectious causes [2]. Now that viral infections appear responsible for some of the malignant effusions (eg, Kaposi sarcoma, multicentric Castleman disease, primary effusion lymphoma), the line between infectious and noninfectious causes is blurred. In addition, much of the data regarding prevalence come from hospitalized patients with advanced HIV infection in the pre-potent antiretroviral therapy (ART) era and may not reflect the post-potent ART experience.

The etiology of the pleural effusion can be established in the majority of patients with AIDS, although a small percentage remains undiagnosed, as is the case in the non-AIDS population [2]. (See "Diagnostic evaluation of the hemodynamically stable adult with a pleural effusion".)

The common causes of pleural effusions in patients with HIV infection are reviewed here. The general approach to the investigation of a pleural effusion is discussed separately. (See "Pleural fluid analysis in adults with a pleural effusion" and "Ultrasound-guided thoracentesis".)

INFECTIOUS CAUSES — In patients with HIV infection, many of the same infectious agents that cause pneumonia should be considered as potential etiologies of pleural effusion (table 1A-B). Among studies of hospitalized patients with a pleural effusion and HIV infection, an infectious etiology was identified in about two-thirds [1-3]. In general, opportunistic infections of the pleura should be considered in patients with a CD4 count below 150 cells per microL, as these patients have an increased risk of opportunistic infection, possibly approaching that of pre-potent antiretroviral therapy (ART) patients.

Bacterial pneumonia — Bacterial pneumonia is the cause of pleural effusion in approximately 30 percent of AIDS in hospitalized patients with AIDS [1-3]. However, in a series of 137 patients with community-acquired pneumonia, HIV-infected patients had a higher rate of parapneumonic effusions than non-HIV-infected patients [5]. With the introduction of highly active retroviral therapy (HART), the incidence of bacterial pneumonia [6] and the complication of parapneumonic effusion have decreased in HIV infected patients [7].

Most of these parapneumonic effusions are small and ipsilateral to the side of the pneumonia. The fluid is exudative and generally nucleated cell counts are low [2]. Empyema is common and, compared to immunocompetent patients, staphylococci were significantly more often isolated from pleural fluid culture in patients with AIDS [5,8].

Streptococcus pneumoniae accounts for more than 50 percent of these parapneumonic effusions; Staphylococcus aureus is the next most common etiologic organism. Less frequent causes include Haemophilus influenzae, Mycoplasma pneumoniae, Legionella, Nocardia asteroides, and Bordetella bronchiseptica [2,9]. Septic pulmonary embolization from right-sided endocarditis in intravenous drug abusers can also result in small pleural effusions.

Treatment of bacterial pneumonia in HIV-infected patients is discussed separately. Uncomplicated parapneumonic effusions generally resolve with treatment of the underlying pneumonia. (See "Bacterial pulmonary infections in patients with HIV" and "Epidemiology, clinical presentation, and diagnostic evaluation of parapneumonic effusion and empyema in adults".)

Pneumocystis jirovecii — Pneumocystis jirovecii pneumonia (PCP) remains an uncommon cause of pleural effusion in HIV-infected patients, with an incidence varying from 3 to 15 percent in hospitalized patients [1,2]. Based on a small case series, there is a suggestion that pleural effusion due to PCP is more common in patients receiving aerosolized pentamidine prophylaxis [10]. In addition to pleural effusion [11-13], PCP also is associated with pleural masses and pneumothorax [1,14-16].

Pleural effusions due to PCP are usually small to moderate in size; can be either unilateral or bilateral; and are serous to bloody in appearance. Pleural fluid acidosis (ie, pH <7.20) has not been reported. Lactic dehydrogenase (LDH) is elevated, with a pleural fluid to serum ratio greater than 1, characterizing the fluid as an exudate. Total protein pleural fluid to serum ratio tends to be less than 0.5, and pleural fluid glucose and amylase are equivalent to serum values [10]. The total nucleated cell count is low, and the differential shows predominantly mononuclear cells.

A definitive diagnosis requires demonstration of the trophozoite or cyst forms of Pneumocystis in the pleural fluid. Grocott silver staining will show the cysts; however, if the pleural fluid has predominantly trophozoite forms, silver stain may not be sufficiently sensitive to establish the diagnosis [11]. In such cases, examination of an air-dried Diff-Quik stained cytospin preparation is ideal [12].

Other markers of Pneumocystis pleural infection are the presence of granular eosinophilic aggregates, numerous vacuolated macrophages, and scattered benign mesothelial cell clusters in the pleural fluid. (See "Epidemiology, clinical presentation, and diagnosis of Pneumocystis pulmonary infection in patients with HIV".)

Neither the treatment nor the prognosis of PCP is altered by the development of a pleural effusion. On the other hand, the development of pneumothorax in these patients is often an ominous occurrence, and usually signals the need for prolonged chest tube drainage.

The treatment of P. jirovecii in HIV-infected patients is reviewed separately. (See "Treatment and prevention of Pneumocystis infection in patients with HIV".)

Mycobacteria — Mycobacterial infection accounts for 8 to 10 percent of pleural effusions in patients with AIDS [1,2]. In most series, Mycobacterium tuberculosis (TB) was found to cause pleuritis more often than Mycobacterium avium complex (MAC); however, one study of 130 patients found the opposite in that 6 percent of HIV related effusions were due to MAC and 1 percent to TB [2,3,7,17]. Other nontuberculous mycobacteria rarely cause pleural disease [18]. A discussion of the diagnosis and treatment of tuberculous pleurisy in HIV-infected patients is provided separately. (See "Tuberculous pleural effusion".)

Pleural TB may be present with or without visible lung parenchymal disease. In a series of 70 patients with HIV infection and pleural TB, the chest radiograph showed an isolated pleural effusion, associated pulmonary parenchymal infiltrate, and a miliary pattern in 51, 37, and 8 percent, respectively [19]. In a case report, a pleural effusion due to MAC infection was large enough to cause hemodynamic compromise [20].

The pleural effusion is virtually always exudative and usually demonstrates neither pleural fluid acidosis nor low glucose levels [21]. The nucleated cell counts in the pleural fluid are lower than those seen in TB pleurisy of HIV-uninfected patients; a predominance of lymphocytes is usually noted. Pleural fluid lymphocytes and CD14 positive macrophages have been shown to be a sanctuary for increased HIV replication in patients with Mycobacterium tuberculosis pleurisy [22]. Molecular testing for Mycobacterium tuberculosis/resistance to rifampicin (MTB/RIF) assay has some value for rapid (ie, within two hours) diagnosis of pleural tuberculosis, especially when performed on pleural tissue. However, the routine use of this assay is not recommended in pleural fluid due to its low diagnostic sensitivity (approximately 45 percent) [23-26].

Other diagnostic tests include:

●The tuberculin skin test (PPD) is positive in 20 to 64 percent of patients [19,27].

●Pleural fluid adenosine deaminase (ADA) is typically greater than 50 U/L in tuberculous pleural effusions, independent of HIV status [28].

●Sputum acid-fast smear is positive more often in HIV-infected patients with advanced immunodeficiency compared to those with relatively preserved CD4 cell counts [19,27]. Sputum induction may be helpful [29].

●The pleural fluid acid-fast smear is positive in less than 15 percent of tuberculous effusions.

●Pleural biopsy for acid-fast stain and pathologic examination for granulomas is helpful with yields better than 50 percent [19,21].

The treatment of TB in HIV-infected patients is discussed separately. (See "Treatment of drug-susceptible pulmonary tuberculosis in nonpregnant adults with HIV infection: Initiation of therapy".)

Immune reconstitution inflammatory syndrome (IRIS) refers to paradoxical worsening of systemic inflammation associated with preexisting infectious processes following the initiation of ART in HIV-infected individuals [30]. This is more likely in patients with a CD4 count under 100 cells per microL and in the setting of initiation of TB treatment [30,31]. Patients may present with worsening constitutional symptoms, pulmonary parenchymal disease, adenopathy, or pleural effusion [31,32]. (See "Immune reconstitution inflammatory syndrome".)

Nocardia — Nocardia asteroides has been identified in about 3 percent of pleural effusions in patients with AIDS [2,33,34]. It is generally associated with more advanced AIDS. The evaluation and treatment of nocardia are discussed separately. (See "Nocardia infections: Epidemiology, clinical manifestations, and diagnosis" and "Treatment of nocardiosis".)

Cryptococcus — Cryptococcus neoformans can occasionally cause pleural effusions in patients with AIDS [35]. In AIDS patients, an isolated pleural effusion may be the initial manifestation of disseminated cryptococcal disease [2,36-39]. C. laurentii, another strain of cryptococcus, has been also identified as a causative agent for pleural effusion in AIDS [40]. (See "Cryptococcus neoformans infection outside the central nervous system".)

In cryptococcal effusions, the pleural fluid is typically a serous exudate with a low nucleated cell count. Pleural fluid culture is diagnostic in 42 percent of cases [41]. Cryptococcal antigen determination in pleural fluid of non-AIDS patients has proved to be a rapid and specific test [42]. Therefore, serum and pleural fluid cryptococcal antigen tests should be performed in all AIDS patients with pleural effusion, since a positive result may allow early treatment and prevent dissemination. Pleural fluid culture is usually positive within a week; however, an incubation time up to two weeks may be necessary before a fluid sample is considered negative for Cryptococcus [38]. Pleural biopsy has been helpful in making a definite diagnosis [38]. Pleural fluid adenosine deaminase concentration has been reported to be falsely elevated in cryptococcal pleural effusion [43]. The pleural effusion resolves with amphotericin B, but may also resolve spontaneously [44].

Other rare causes for pleural effusion in AIDS worldwide include paracoccidioidomycosis [45], filariasis [46] and actinomycosis [47]. (See "Lymphatic filariasis: Epidemiology, clinical manifestations, and diagnosis" and "Lymphatic filariasis: Treatment and prevention" and "Clinical manifestations and diagnosis of acute/subacute paracoccidioidomycosis".)

Histoplasma capsulatum — Disseminated histoplasmosis is an important opportunistic infection in AIDS; however, Histoplasma capsulatum is rarely a cause of pleural effusion in patients with or without AIDS [41,48]. Histoplasma pleural effusion may present in association with dissemination, contiguous spread, hypersensitivity pleuritis, or active pericarditis. (See "Diagnosis and treatment of disseminated histoplasmosis in patients without HIV".)

The effusion is serous or serosanguineous, with a low number of nucleated cells. A cytocentrifuged pleural fluid specimen stained with Wright-Giemsa stain will demonstrate intracellular yeast typical of Histoplasma capsulatum. The pleural effusion usually resolves without treatment; however, resolution may be slow, over several months [41].

Toxoplasma gondii — Toxoplasma gondii is a rare cause of pleural effusion [49]. Pleural effusion may be unilateral or bilateral. A cytocentrifuged Wright-Giemsa-stained specimen of pleural fluid will demonstrate T. gondii tachyzoites within neutrophils. Tachyzoites can also be found in bronchoalveolar lavage fluid when there is concomitant pulmonary parenchymal disease.

The treatment of pulmonary T. gondii infection in HIV-infection is discussed separately. (See "Toxoplasmosis in patients with HIV".)

Talaromycosis — T marneffei has been reported as a cause of pleural effusion from Southeast Asia [50]. In a comparative observational study, pleural effusion was observed in 3 of the 19 patients with HIV. Pleural fluid chemistry and cell count data were not available in this study. However, the pleural fluid data from non-HIV population has described the effusion as exudative with neutrophils as the predominant cell type [51].

Leishmaniasis — Visceral leishmaniasis is a common infection in patients with AIDS who live in endemic areas. Rare cases of pleural effusion have been reported [52]. (See "Visceral leishmaniasis: Epidemiology and control" and "Visceral leishmaniasis: Clinical manifestations and diagnosis" and "Visceral leishmaniasis: Treatment".)

MALIGNANT EFFUSIONS — HIV-infected patients have an increased propensity to develop malignancies that cause pleural effusions, particularly if their CD4 count is below 200/mm³ (table 1A-B). Some of these (eg, Kaposi sarcoma [KS], multicentric Castleman disease [MCD], primary effusion lymphoma [PEL]) are related to infections with human herpesvirus 8 (HHV-8) and/or Epstein Barr virus (EBV) [53]. Although antiretroviral therapy (ART) seems to have reduced the incidence of KS, the incidence of MCD and PEL remains unchanged [54]. Patients with AIDS can also develop other cancers that may cause pleural effusions; these are reviewed separately [3]. (See "HIV infection and malignancy: Epidemiology and pathogenesis".)

Kaposi sarcoma — Pleural effusion arises in 36 to 60 percent of patients with KS [55,56]. The onset of pleural effusion is preceded by the appearance of cutaneous KS lesions by several months (picture 1A-C), and usually by a fall in CD4 count to below 150 cells per microL and an increase in viral load to above 10,000 copies per mL [57]. Whereas about 45 percent of subjects with skin lesions tend to have pleuro-pulmonary lesions, the majority of patients with pulmonary KS will have demonstrable skin lesions [54]. Pleuritic chest pain is absent, perhaps because the visceral pleura is preferentially involved, not the parietal pleura. (See "AIDS-related Kaposi sarcoma: Staging and treatment" and "Pulmonary involvement in AIDS-related Kaposi sarcoma".)

In addition to the pleural effusion, the chest radiograph typically shows unilateral or bilateral interstitial and/or alveolar infiltrates; hilar or mediastinal adenopathy is present in 25 to 60 percent of patients with pleuropulmonary Kaposi sarcoma [55,56]. Pleural effusions are predominantly bilateral and are large to massive in size (image 1). Pleural fluid was serosanguineous and free flowing in a majority of patients; hemorrhagic effusions are also seen.

In a review of 105 postmortem examinations of patients with KS, thoracic findings included bilateral hemorrhagic pleural effusions, purplish lesions of KS on the visceral pleura, and infiltration of lung parenchymal lymphatics and mediastinal lymph nodes [55]. Effusions are typically serosanguineous due to hemorrhage from highly vascular lesions of the visceral pleura. Occasionally, a characteristic parietal pleural lesion can be observed during thoracoscopy (picture 2) [58].

Pleural fluid characteristics are as follows:

●The nucleated cell counts are generally low, and there is mononuclear cell predominance

●The fluid is exudative

●The LDH varies between 111 and 330 IU/L

●The pleural fluid glucose concentration is greater than 60 mg/dL (3.3 mmol/L)

●Cytologic examination of the pleural fluid is unrewarding

Chylous effusions have been increasingly recognized in KS, but the hemorrhagic nature of the effusion may mask the typical appearance of a chylothorax [55,59]. A pleural fluid triglyceride concentration more than 110 mg/dL will suggest the presence of a chylous effusion; the presence of chylomicrons will confirm the diagnosis. (See "Etiology, clinical presentation, and diagnosis of chylothorax".)

The diagnosis of pleural KS cannot be made by percutaneous pleural biopsy, because the lesions are rarely found on the parietal pleura. Although open lung biopsy establishes the diagnosis in 75 percent of cases, the high risk/benefit ratio makes this procedure of limited value [60]. The finding of characteristic red endobronchial lesions on bronchoscopy essentially establishes the diagnosis of pleural Kaposi sarcoma in the proper clinical setting, such as a patient with bilateral hemorrhagic effusions [60]. Occasionally, thoracoscopy can demonstrate a typical KS lesion (picture 2) [58]. While human herpes virus (HHV)-8 viral load in bronchoalveolar lavage (BAL) fluid was shown to be elevated in tracheobronchial KS, it has not been tested in pleural KS [61].

Reports of therapy directed to the pleural space are few. Effusions from KS often need drainage for symptomatic relief. Tetracycline pleurodesis was ineffective in a limited number of patients [55]. Radiation therapy has been used as a palliative measure [62]. A single case report indicated resolution of pleural and pericardial effusion following therapy with intravenous pegylated liposomal doxorubicin, pleural drainage, and one session of whole lung irradiation [63]. Other potential newer therapies that are being tried include: interferon-alpha, mTOR inhibitors such as rapamycin, or angiogenesis inhibitors such as lenalidomide, bevacizumab, sunitinib, or sorafenib [64,65]. Anecdotally, some patients may respond to treatment of KS with antiretroviral therapy. The management of malignant pleural effusions and treatment of KS are reviewed separately. (See "AIDS-related Kaposi sarcoma: Staging and treatment" and "Management of malignant pleural effusions".)

About two-thirds of patients with KS associated with the immune reconstitution inflammatory syndrome (IRIS) will have pleural effusions, following initiation of ART [66,67]. The management of IRIS is described separately. (See "Immune reconstitution inflammatory syndrome".)

Multicentric Castleman disease — MCD is a lymphoproliferative disorder that is highly associated with HIV and HHV-8 infection. Typical pulmonary findings are reticular or nodular pulmonary parenchymal opacities, mediastinal adenopathy, and, less commonly, bilateral pleural effusions [68]. The diagnosis is usually made on a lymph node biopsy. (See "HHV-8/KSHV-associated multicentric Castleman disease".)

Lymphoma — B-cell, non-Hodgkin lymphoma (NHL) arises in 5 to 10 percent of patients with AIDS. The reported prevalence of pleural effusion due to lymphoma varies from 1 to 16 percent [3,17,69]. However, an autopsy series documented pleural effusion in 65 percent of those with thoracic involvement with lymphoma and in about 25 percent of all patients with NHL [70]. (See "HIV-related lymphomas: Clinical manifestations and diagnosis".)

Pulmonary involvement with non-Hodgkin lymphoma is more commonly seen in patients with advanced HIV infection [70]. Peripheral lymphadenopathy, hepatosplenomegaly, and elevated serum lactic dehydrogenase (LDH) were common. Pleural effusions were associated with pulmonary nodules, areas of consolidation, or masses; about 60 percent had associated thoracic lymphadenopathy. Pleural effusions caused by NHL may have a mass effect, whereas effusions from primary effusion lymphoma do not. (See 'Primary effusion lymphoma' below.)

When NHL involves the pleura, effusions tend to be large, serosanguineous, lymphocyte-predominant exudates with a normal glucose, and a high LDH concentration [70]. Pleural fluid cytology has been reported to be diagnostic in about 60 to 75 percent of cases [70,71]. Pleural biopsy also had a high diagnostic yield [70]; flow cytometric examination of pleural fluid has also been useful in establishing the diagnosis [72].

The treatment of NHL in HIV infected patients is reviewed separately. (See "HIV-related lymphomas: Treatment of systemic lymphoma".)

Primary effusion lymphoma — Primary effusion lymphoma (PEL) is a high-grade NHL of B-cell origin that is predominantly found in HIV-seropositive gay men between the ages of 30 and 50 years [73-77]. PEL is strongly associated with HHV-8 infection. Preexisting KS has been reported in 33 to 70 percent of patients with PEL [78]. Increasingly co-infection with EBV has been reported in PEL [79]. (See "Primary effusion lymphoma".)

PEL cell lines have been shown to generate excessive vascular endothelial growth factor (VEGF)/vascular permeability factor (VPF); VEGF/VPF may contribute to excessive fluid formation in the pleural space, pericardial sac, and abdomen [80]. The absence of lymphadenopathy suggests that central lymphatic obstruction by lymphoma is not the mechanism for fluid formation.

Chest computed tomography (CT) shows pleural effusion with adjacent parietal pleural thickening in the absence of lung parenchymal lesions or mediastinal lymphadenopathy. Pericardial effusion and ascites are other common radiological findings. Peripheral lymphadenopathy and bone marrow involvement are unusual findings [81].

The diagnosis of PEL is usually made by the cytologic examination of pleural fluid (picture 3). Pleural fluid immunocytochemistry shows malignant cells that are often positive for CD45 and CD30, and variable for epithelial membrane antigen [82,83]. The nuclei usually stain positive for HHV-8; the most common method for detecting HHV-8 positivity is immunohistochemical staining for the latent viral gene product known as latency-associated nuclear antigen (LANA-1) [84].

The treatment of PEL is discussed separately (see "Primary effusion lymphoma"). Median survival with treatment in PEL was 6.2 months. The predictors of poor outcome were poor performance status and having not received ART prior to diagnosis [85].

Lung cancer — The incidence of lung cancer is significantly increased in patients with HIV infection, and malignant pleural effusions are a relatively common sequela of lung cancer in patients with HIV/AIDS. (See "HIV infection and malignancy: Management considerations", section on 'Lung cancer'.)

OTHER NONINFECTIOUS CAUSES — Other noninfectious disorders that may cause pleural effusions in HIV-infected patients include hypoalbuminemia, heart failure, uremia, and pulmonary embolism (table 1A-B) [1].

Hypoalbuminemia — Hypoalbuminemia has been encountered as a cause for pleural effusion in 7 to 15 percent of hospitalized patients with AIDS [1,2]. Because hypoalbuminemia is a recognized problem in progressive AIDS, hypoalbuminemic pleural effusions may become more prevalent as the infectious complications of AIDS are managed more effectively. Hypoalbuminemic effusions are small to moderate in size, bilateral, and transudative. The nucleated cell count is low and mononuclear cells predominate.

Heart failure — Pleural effusion secondary to cardiac failure has been noted in hospitalized patients with AIDS [2]. In this setting, dilated cardiomyopathy is usually caused by infectious myocarditis; toxoplasmosis, cytomegalovirus, and HIV have all been implicated [86]. Pleural effusions are small, unilateral or bilateral, and transudative. The effusions generally are clinically unimportant and resolve with treatment of the heart failure. (See "Cardiac and vascular disease in patients with HIV", section on 'Myocardial disease'.)

DIAGNOSIS — The diagnostic evaluation of pleural effusions in HIV-infected patients follows a similar algorithm to that for HIV-noninfected patients; however certain additional tests may be important (table 1A-B) (see "Pleural fluid analysis in adults with a pleural effusion"). If these tests are not diagnostic, further investigation should follow the steps for an undiagnosed pleural effusion. (See "Diagnostic evaluation of the hemodynamically stable adult with a pleural effusion".)

SUMMARY AND RECOMMENDATIONS

●Introduction – The majority of pleural effusions in patients with AIDS are secondary to infection; however, about a third are due to noninfectious causes. (See 'Introduction' above.)

●Infectious causes – Infectious causes include the following (see 'Infectious causes' above):

•Bacteria – Bacterial pneumonia is the single most common cause of pleural effusion in AIDS, accounting for approximately 30 percent of effusions in hospitalized AIDS patients. (See 'Bacterial pneumonia' above.)

•Pneumocystsis – Pleural effusion is an unusual manifestation of Pneumocystis jirovecii (PCP) infection in HIV-infected patients. PCP pleural effusions are usually small to moderate in size, either unilateral or bilateral, and serous to bloody in appearance. Trophozoite or cyst forms of P. jirovecii are found on Diff-Quik or silver stains of pleural fluid, respectively. (See 'Pneumocystis jirovecii' above.)

•Mycobacteria – Mycobacterial infection accounts for 8 to 10 percent of pleural effusions in patients with AIDS; depending on the patient population, M. tuberculosis or M. avium complex may be more common. Acid-fast stain is positive in 15 percent of pleural fluid samples and 69 percent of pleural biopsy specimens; granulomas are present in up to 88 percent of the biopsies. Molecular testing for Mycobacterium tuberculosis/resistance to rifampicin (MTB/RIF) assay has some value for rapid diagnosis of pleural tuberculosis on pleural tissue but not in pleural fluid where its diagnostic sensitivity is low. (See 'Mycobacteria' above.)

•Others – Other opportunistic infections may occasionally cause pleural effusions in HIV-infected patients, including Cryptococcus, Histoplasmosis, and Toxoplasma.

●Malignant effusions – Malignant causes include the following (see 'Malignant effusions' above):

•Kaposi sarcoma – Pleural effusions are common among patients with Kaposi sarcoma (KS); the fluid is usually serosanguineous to hemorrhagic. Virtually all patients have skin lesions of KS and many have endobronchial KS. Pleural cytology and pleural biopsy are negative. Pleuroscopy can detect typical pleural KS lesions. (See 'Kaposi sarcoma' above.)

•Lymphoma – When non-Hodgkin lymphoma (NHL) involves the pleura, concomitant pulmonary nodules, masses, or consolidation are also present on CT; about 60 percent of patients have mediastinal lymphadenopathy. Pleural fluid cytology and flow cytometry, pleural biopsy, and transbronchial biopsy of lung lesions have high diagnostic yields. (See 'Lymphoma' above.)

•Primary effusion lymphoma (PEL) – PEL is a high-grade NHL of B-cell origin. Unlike pleural effusions from MCD and other lymphomas, there is no associated lymphadenopathy. The diagnosis of PEL is usually made by the cytologic examination of pleural fluid with immunohistochemical staining for HHV-8, such as the viral gene product, latency-associated nuclear antigen (LANA-1). The median survival with treatment has been reported to vary from three to nine months. (See 'Primary effusion lymphoma' above.)

●Other noninfectious etiologies – Other noninfectious disorders that may cause pleural effusions in HIV-infected patients include hypoalbuminemia, heart failure, and pulmonary embolism. (See 'Other noninfectious causes' above.)

●Diagnostic evaluation – The diagnostic evaluation of pleural effusions in HIV-infected patients follows a similar algorithm to that for HIV-noninfected patients; however, certain additional tests may be important. Addition of chest CT scan improves the recognition of undiagnosed pleural effusion and parenchymal abnormalities in patients with AIDS being evaluated for abnormal chest radiograph. (See 'Diagnosis' above and "Pleural fluid analysis in adults with a pleural effusion".)